研究拱坝结构抗震性能及破坏形态的重要手段为拱坝振动台动力模型试验，试验配制一种以重晶砂、重晶石粉和水泥为主要成分的模型材料，结合选用的振动台所具备的条件及模型材料的力学性能，设计拱坝振动台试验模型，并测试和获取其在不同地震荷载下的动力响应。选择某一工况进行分析，该工况下结构开始出现明显裂缝损伤并逐渐发展。通过分析该工况下结构动应变响应，分析其损伤出现的时间，由此将该工况分为4个时段。通过计算分析不同时段多个加速度响应测点的加速度放大系数，并引入模态参数辨识的数值子空间辨识算法和ARX模型辨识法，计算不同时段+拱坝模型的自振频率和阻尼比，并将其与损伤发展情况相对比，以分析拱坝模型损伤发展情况及其结构动力特性。研究结果可为拱坝振动台试验制作和结果分析提供价值性参考，并可为拱坝结构动力特性参数识别和损伤诊断方法提供验证素材。

The dynamic model test of an arch dam on a shaking table is an important way to study its seismic performance and failure mode. A shaking table damage model test of the arch dam was carried out. A model material primarily composed of barite sand, barite powder, and cement was made. The model material had material possesses characteristics of high density, low dynamic elastic modulus, and appropriate strength, with its average density being 3,175 kg/m3 , average dynamic elastic modulus being 1.05 GPa and average tensile strength being 68.7 kPa. Combining the conditions provided by the selected shaking table and the mechanical properties of the used model material, the arch dam model was designed with four parts: dam body, foundation, dam shoulder, and boundary wall. The height of the dam body, the maximum width of the dam crest, the thickness of the dam crest, and the thickness at the base of the dam body were 1.5 m, 1.0 m, 6 cm, and 11 cm, respectively. The total mass of the arch dam model was 2.17 t.Pressure-resistant accelerometers and resistance strain gauges were arranged at different parts and elevations of the arch dam model to obtain dynamic responses. The structural dynamic responses of the model under different earthquake simulation loads were tested. A specific working condition (Time period was 32.0 s) of the arch dam model was selected for analysis, in which the structure began to exhibit evident damage with visible cracks in arch crown and gradually progressed. By analyzing the structural dynamic strain responses under following working condition, the time of the damage occurrence was determined at t=5.8 s, and the condition was divided into four periods. Multiple acceleration response measurement points on the arch crown beam were selected for analysis. The acceleration amplification factors at the acceleration measurement points near the arch crown showed a noticeable and gradually increasing trend in four time periods, consistent with the characteristic of significant cracks appearing in the middle part of the arch crown.The subspace numerical algorithm for modal parameter identification (N4SID method) and the ARX model identification method were introduced. Based on the collected dynamic response measurements, the identification of natural frequencies and damping ratios of the arch dam model structure were conducted for four different periods by the introduced two methods. It was observed that both methods yielded small errors in identifying natural frequencies at each order, while there was a significant difference in the identified damping ratios. This discrepancy was primarily attributed to the high complexity of the damping characteristics of the arch dam structure. Comparing the identified values of natural frequencies and damping ratios at each order for the four time periods, it was found that the natural frequencies gradually decreased with the development of structural damage, while the damping ratios increased with the development of structural damage. The identification results were consistent with the strain response process and the actual damage results of the arch dam model, validating the effectiveness of the dynamic characteristic parameter identification. The results provided a valuable reference for the preparation and analysis of arch dam shaking table tests. Additionally, this study can also serve as validation material for the identification of dynamic characteristic parameters and damage diagnosis methods in arch dam structures.